Current content distribution techniques include the use of rate adaptive encoded content. A device processing or consuming rate adaptive encoded content can vary the bitrate of the content (e.g., video) as network conditions change.
Initial resolutions and/or frame-rates used during playback might be poor as the device typically will not know what bitrate can be expected on the network. In order to minimize perceived or actual start-up time, the device may initially select a lowest quality stream or resolution/frame rate and then “ramp up” in terms of selecting a higher quality encoding. The “ramp up” may proceed on an iterative basis or in lock-step until the highest quality encoding that the network can support is selected.
There are a number of drawbacks associated with conventional solutions that are identified and addressed in this disclosure. For example, the provisioning of higher quality encoded content may be delayed. For example, the delay may be a result of overly conservative bandwidth estimates. Storage under runs may result in content temporarily being unavailable. Content stored in a buffer or memory might not be improved or enhanced for secondary or subsequent consumption, such as supporting a secondary viewing of video via a rewind command or function, without discarding the initially saved/buffered segment and re-downloading the segment at higher quality.